Enhanced Handover Management in 5G Networks using Machine Learning in Fading Scenarios

This study explores the use of machine learning (ML) to improve initial handover performance for 5G networks by analysing fading-affected scenarios. In this study, we assess the ability of ML models to predict optimal handover decisions compared with conventional methods. In this study, simulation results show that our ML-based methods can reduce the handover inaccuracy by 50%, decrease latency up to 10 times and stabilize network performance. We accompany our assessment with a series of graphical analyses to show how ML algorithms change their behaviour dynamically in response to deteriorating link intensity due user mobility and increasing h/o frequency, thereby degrading the network performance. The results clearly illustrate that ML enabled handover optimization is significantly more effective than traditional methods in accommodating fading-channel dynamics, thereby improving both the network performance and user experience of future generation wireless networks.